The Right Tool For The Job and ROI

27 05 2010

AppleInsider had an article on May 18, 2010 which was titled “Apple Scaling Final Cut Studio Apps to fit prosumers” which generated a ton of blogosphere panic. Even I was caught up in the rumor mongering here, reacting to a post I’d read on Twitter and then, after reading the AI piece, tweeting about it myself. Phillip Hodgetts had a very intelligent post on his blog last week that used a historical approach to take the AppleInsider piece apart, rebutting nearly everything that the article said. Larry Jordan followed up with another article which also took pains to point out why that original piece was Dead Wrong.

But, in doing so, he made another excellent point.

For me, this is the key point — as editors our job is to tell stories visually. The tools we have today do a really great job of helping us put food on the table and pay the rent.

The emphasis is mine, by the way.

Now, I’d be the last one to paraphrase Larry (though I will be doing a bunch of that in a vidcast with him which will start in mid-June — more details on that to come), but let me try. What I think was so cogent about Larry’s comment is this: We only need enough tools to do the best job we can.

Of course, there’s a lot to pick apart in that statement. We were fine working on 35mm and 16mm film, drawing diagonal grease pencil lines down the middle of the film to indicate dissolves. But then videotape editing came along and, soon, we were able to actually see that dissolve. Very quickly, those diagonal lines were not “doing the best job” anymore.

Then there’s the reality that one editor’s “need” is another one’s “nice to have but I don’t care.” New tools in Avid’s Media Composer make displaying 3D footage must easier, but most everyone I know doesn’t work in 3D so (for now) we won’t care about it.

But those issues aside, the truth of that statement is strong. It’s not as important for us to have access to every tool out there, as it is to have the right tool. Until very recently, many feature films were edited on a very old version of Avid’s Media Composer hardware and software because that version of the program was stable, worked beautifully and gave editors everything they needed. Of course, with the advent of HD and visual effects, you can’t say that anymore, except if your job only involves straightforward SD editing. Then the urge to upgrade just isn’t there. Businesses call it ROI (“return on investment”) and the equation holds true in editing as well. Will we make or save as much money upgrading to a new tool as it will take to buy it, install it and (most importantly) learn it?

As the world changes, our editing tools must change of course. But the inverse is not necessarily true; as our editing tools change, the world doesn’t have to change as well. If something works really well in version 4.0 or in version 6, why should we upgrade to 5.0 or 7?

Incorporating new technology into our own work lives can be fraught with peril and we’ll only jump at the changes that make sense. How can we determine what makes sense without reflexively avoiding something just because it’s a change, or darting to every new bell and whistle just because it is new? Good question. We deal with that all the time.

Recently, I’ve been playing with two tools that are designed to make editing life more sensible and I’ve now incorporated them into my own editing life. In each case, I got something more by the change, than I had to put out in order to make that change. That is real life ROI.

I first saw PluralEyes back at NAB in 2009, where it was stuck all the way at a side wall. The way it was pitched to me got my juices excited — this is a tool for editors (FCP only at the time, it has now expanded to Premiere and Vegas; where is Media Composer???) that will automatically sync takes from different cameras that were shot at the same time and have matching audio. This seemed to be a godsend editors of music videos or events (think speeches or weddings) that are captured using multiple cameras. Six cameras capturing a speech can be easily sunk up to each other, even if the audio is of varying quality. Editors who have to sync multiple takes of a musical performance that was shot to a common playback will also benefit from this.

What a cool idea, right? I can hear editors all over the world counting up the amount of time that they will be saving in syncing up footage. In the “old days” this would have involved finding common points between each and every take (a verse where the band sang the word “Killer”, for instance — hard consonants like “K” are useful in finding sync), mark a sync point at those points in all of the takes, and combine the takes into one multicamera clip. This was pretty reliable but was incredibly time consuming and prone to error, especially if the person doing the syncing had to make sure that he/she wasn’t using that same word, but from different verses. In addition, at times the audio on an individual camera might not have been at the same level or quality as another camera, making it harder find the exact match by listening or looking at the audio waveforms in our NLEs.

So, PluralEyes could be a great timesaver but in order to do that, it has to require less work to set up than we benefit by using it. As examples, Avid’s ScriptSync used to take too much of my editing time to set up and so I never used it. Once they put voice recognition into it, it became a very usable tool and I now love it. On the other hand, I’m still waiting for Adobe’s Transcription tool to get to a usable state — right now I get around 50% accuracy, which creates more work fixing a transcription than I’ve saved by doing it automatically in the first place — Scott Simmons has a great review of it in his Editblog.

So, was PluralEyes helpful? Does it pass that test?

Way yes!! It can’t sync everything, but it does a great job of finding the sync points between takes, even if one of the clips is only a partial subclip from waaaaay down in a take. It does a remarkable, though not flawless, job in matching audio recorded at different levels and echo. I was able to effortlessly sync two cameras with direct feed audio, up to one that was using the camera mic, with all of its attendant room echo and noise. In the one or two cases where, for no known reason, it couldn’t sync up a track, it created a separate FCP timeline with those clips on it. This made it easy to see what wasn’t automatically sunk up so I was able to hand-sync those pieces. Synching two or three pieces, rather than thirty, is a huge time saving and so PluralEyes deserved to be in my editing tool chest.

It was the Right Tool for that very limited job and, even at $149, that was way worth it (Honesty Policy: Singular sent me a review copy of PluralEyes, so I didn’t pay that $149. But that doesn’t change my feeling about its worth.) I don’t know what your pay scale is, but if you use this application for three jobs and it saves you two hours in each, that’s about $25 an hour. If you’re not charging at least that for your time, you are either a student or starving or both. One key to this program’s success is its laser beam focus on one thing — help editors sync audio takes together quickly. That’s it. Priced accordingly, it’s a no-brainer for anyone who needs that one thing.

As an aside, Larry Jordan mentioned in his May 20, 2010 Digital Production Buzz podcast, that he has more editing applications on his computer than you can “shake a stick at”. (I’m not sure why you’d want a shake a stick at a computer — I often shake my fists, but that’s different.) He went on to say that he used different ones because not every NLE is good as another at specific things. I got to thinking about that. I used Media Composer a lot for my editing, but I absolutely hate their Titles creation tool — both Marquee and AvidFX/Boris — so I usually bop over to Motion to create lower thirds and the like and then import those files into my Avid machine. The right tool for the job. This is another example of creating a focus on single tasks. When I want to teach students how to create a simple DVD I’d rather use iDVD than DVD Studio Pro (even in it’s simple mode) because it’s Stupid Easy. But it’s phenomenally awful to do anything more complicated. For that I use DVD Studio Pro.

I apologize here for my total lack of knowledge of most Adobe products. I’ve been quite impressed by their improvements in the last few years, but my main body of knowledge still revolves around the NLEs that we use most here in the US — primarily the Media Composer and Final Cut.

Sorenson 360Another tool that I’ve been testing on and off for several months is something called Sorenson 360, which makes it much easier to upload videos that I’ve created for viewing and approval by my producing and directing collaborators. It will come as no surprise to those of you who have been reading this blog for a while that I am a strong proponent of long distance collaboration. I believe that, for editors of the future to be successful, we are going to have to be working with clients all over the world, often many of them at the same time. The feature I’m cutting now has me sitting in front of my computer in Los Angeles, the director is in Rhode Island and the producer is in Massachusetts. We need to be able to easily show each other sequences without flying all over the U.S. To that end, a number of cloud-based review and approval sites have been born on the web. They make compressing, commenting and approving much easier.

Sorenson 360 does all of that to great degree. Like any good compression tool, Sorenson Squeeze can take a while to efficiently and decently compress your films. For a 2 minute trailer that I recently created for that feature I mentioned, it took over an hour. For a documentary that I’m editing on Global Rivers, I had to create a 12 minute excerpt reel. The compression on that sequence, which was originally shot in HD/P2 format, took at least three hours — I left it after about 50 minutes and let it work overnight. When it was done, I had the site send me and my producers an email message that the upload was ready for them and gave them the password. It could have also sent us a text message as well.

Now, as anyone who has ever done any compression can tell you, finding the right compression settings is never as easy as they tell you. I’m okay at this, but I never can find the proper settings for quality, size and platform right out of the gate. Most compression programs give you a number of presets for each use but I find that these are no more than starting points. I am continually tweaking the settings for optimal image quality and web playability. Of course, once you determine the best setting for a particular project you should save it in a preset so you can use it all the time without the need to experiment each and every time (and I usually create a preset or two for each project I do — compression seems to be that finicky).

So, Sorenson Squeeze does all of that, as does Compressor. But Sorenson also provides a direct connection to its Content Delivery Network — the aformentioned Sorenson360 — as well as the notifications that streamline the approval process. It also gives me some rudimentary metrics — such as how many views each video received as well as the viewing duration for each video. This is great for web videos so you can basically tell where a viewer stopped watching your show (I find that the average viewer often dumps out of a video part way through — this way you can find out a bit of the “why”).

So, is this a tool that you need? And is it a tool that’s worth the cost (after a year of the free service that comes with Sorenson 6, the costs “start at $99″ and, yes, their website is that opaque about the costs saying that it’s “pay-as-you-go”)? Well, it depends on what you need it for. Brightcove, a leader in the CDN space (also acronymed the “ODN space” — Online Delivery Network), already provides pretty strong streaming in a variety of platforms with a full set of the statistics necessary for advertisers and sponsors. Can Sorenson deliver the same goods? Their prices range from the same $99 per month (50 videos and 40GB of bandwidth) to $5oo (for 500 videos and 250GB bandwidth).

I have to say that I’m not a Brightcove user so I don’t know the answer to that question. The real question is whether I’d reup with Sorenson 360 when my free one-year is up, and that is also a decision based on my own needs. I don’t create so many videos per month that $1200/year is worth it for me. But if you’re a video professional who finds him or herself increasingly working over distances this also might be the right tool for the job. I love its integration with Sorenson Squeeze (my compressor of choice). I love that I can drop a timecode window on top of my video in Squeeze to provide my producers with an easy way to key their notes to a specific spot in the video. I like the RTMP streaming which enables viewers to easily start a video from any point within the stream, rather than start at the top. I don’t like the fact that there are presently only two real formats for display — H.264 or Flash. I’d like some HTML5 capabilities as well. But it’s a great tool; well thought out and (with the recent upgrade to Version 2) becoming increasingly more sophisticated.

To see the example of how I used this tool on the Global Rivers documentary, you can temporarily check it out at my Sorenson360 site. I output this 12 minute trailer to a Quicktime movie, compressed it in Sorenson 6 and uploaded it to that site behind a password which, in this case, is “globalrivers“.

But, for many people, these applications could be another example of The Right Tool. Would it be really cool if we could get all of this in Final Cut or Media Composer? Maybe. Would it be awesome to be able to create Edit Lists or Film Cut Lists right in our NLE (the way we used to in Media Composer) without having to jump out to a separate program? Again — maybe.

Larry Jordan’s point is well taken. Not every tool needs to do everything. In fact, at a certain point, a tool that does everything is going to resemble Microsoft Word, where most users don’t take advantage of 95% of what the program can do, but it loads incredibly slow nonetheless because Microsoft is putting everything in the tool. Every NLE is going to need just the right tools to let the editors do their job, and no more. The real trick, with so many different editors out there, is figuring out just what the bulk of our editors need, and then give them The Right Tools to do that.

[PluralEyes disclaimer added – June 2, 2010]



6 responses to “The Right Tool For The Job and ROI”

27 05 2010
Evan (00:58:07) :

One question I’ve always had about CDNs is whether it’s a liability for me to upload a sequence of a feature I’m cutting, even if it’s private and shared only between me and the director/producer/etc. I’ve always erred on the side of extreme caution and uploaded only to my own servers that I have full control over, but of course that doesn’t always provide for the best results. (For example, my server doesn’t support RTMS, so if I encode a FLV I can’t randomly seek to the middle of a clip.)

I guess I just worry that buried in the fine print of the license agreement for a CDN is a license transfer of my material, temporary or otherwise, which I am usually not in a position to give. On someone else’s servers I also can’t guarantee that the server actually deletes a clip I tell it to, and so all that liability if something goes wrong is usually enough to scare me away from 3rd-party companies’ services and servers.

What are your thoughts on this?

27 05 2010
Evan (00:59:19) :

er, I meant RTMP :)

28 05 2010
Mark Underwood (07:56:26) :

Ive thought for a long time that it would be nice for an NLE to do just that… Just cuts. Maybe a few transitions and basic titling. It seems to me (and bare in mind that I havent used avid) there is just oo much stuff in NLE’1 these days. All the filters and effects etc. You should do all of that somewhere else. Just let me get the footage, cut it, and then allow easy output of those cuts. Save me the processor cycle and ram for playback.

I wonder with the opesourceing of Editshare at NAB if I might be able to finally get this. Here’s hoping :)

28 05 2010
Mark Underwood (07:59:47) :

I Meant Lightworks, thats OWNED by editshare.

Btw anybody used lightworks? thoughts opinions? I think this was one of the most under reported events of NAB. Although that might be because im an editor and Programmer, so im a bit biased.

28 05 2010
Norman (16:46:51) :

Lightworks was the first digital editor I ever used, way back when it was privately owned by a group of crazy filmmakers/developers. I always loved the interface (remember that, back then, not even Avid edited on the timeline — it was just a display) but it was an SOB for media management and suffered from it’s tie to MS-DOS. Eventually, Avid overtook it and a number of the other NLEs that I was playing with — Montage, Ediflex for instance.

A simple, open source, donationware NLE would be kinda awesome. Is that where you think Lightworks should go? It certainly can be more than cuts-only, but where would you stop it? What would you leave out and what would you keep? In a world where iMovie can do so much, the bar for a top free editor is pretty high.

8 06 2010
Mark Underwood (02:06:54) :

I’m glad to hear someone who has used lightworks give feedback and impressions about it, Ive had a hard time getting any info about the actual usage of it.

The line to be drawn for features in software is always a hard one. I think it would be something along the lines of “Does this help you cut?” and if it doesnt, it aint in. Ideally it should just almost be for assembly. SO its the start point and possibly the end point. Any vfx colouring, sound etc, should be done somewhere else, and the editing software should try to be a central point for all information, rather than modifying that information. Im trying to come up with an analogy to do with water spliting apart and then coming back together, but cant come up with an elegant one.

Leave a comment

You can use these tags : <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>